Component Data Whitespace for Sweden: 7,424 Cells in 5 of 8 Walsin Families, and 25 Values That Split in Two

Component data whitespace is invisible and it is not evenly spread. We checked every cell in the 120 Walsin export files that hold parts: 7,424 cells in 5 of the 8 families carry a stray space, tab or line break, and 25 values split into two groups because of it. Trim before you compare.

Venatech Limited, an authorised Walsin distributor in Hong Kong, holds all 203 parametric export files published by Walsin Technology (Walsin), of which 120 contain parts. This post is written from a character-level read of every cell in those 120 files: leading and trailing spaces, runs of two or more spaces, tabs and line breaks, counted per family. A spreadsheet shows none of this. A group-by, a join or an exact-match lookup in an ERP import sees all of it. Every part number below is real.

Which of the eight Walsin families carry hidden whitespace, and how much?

Five of eight, and the density ranges from one part in 450 to three parts in four. We scanned every column of every non-empty file and de-duplicated by part number and column, because the exports repeat parts across combined and per-series files. 7,424 distinct cells in 6,682 part numbers hold whitespace that is not a single space between words. The radio-frequency family is by far the worst, with 353 of 457 parts affected.

We have not seen this counted anywhere. The Part Number column is clean in every family: no leading, trailing or doubled space in any of the 359,000-odd part numbers. The defects live in the descriptive columns, the ones a selector groups on. The embedded single space inside many chip resistor part numbers is a different defect, covered in why Walsin part numbers do not paste into your BOM. We did not count it here.

Which three families are completely clean?

Thin film, disc capacitor and varistor, which together hold 95,222 part numbers. We ran the same scan on every column of those files and it returned zero cells. That is 93,914 thin film, 929 disc capacitor and 379 varistor part numbers with not one leading space, trailing space, doubled space, tab or line break in any column. Thin film is the second-largest family in the catalogue, so the clean result covers real volume.

The clean files can be checked by anyone. Their cells are short and single-valued, such as 1.00Ω in the thin film files or +/-10% in Walsin_MOV_ALL.xlsx. In the other five families the defects sit in cells that hold more than one value, such as the radio-frequency bands below, or in series names. In the clean three, a string comparison behaves as a script expects.

Which kinds of whitespace defect are there, and which is commonest?

Doubled spaces, by a wide margin. We classified every defective cell, and a cell can carry more than one class, so the totals below overlap by 28. Doubled spaces account for 5,620 cells, trailing spaces for 1,044, line breaks for 784, leading spaces for 3 and tabs for 1. Chip resistors supply nearly all the doubled spaces: 5,476 of them are one column in one file.

That file is Thick_Film_Chip_Resistor_Array__1of1.xlsx, the thick film chip resistor array export. Its Series cell reads Thick Film Chip Resistor Array 0402*2, Convex Type with two spaces after the word Array, on all 5,476 array parts. No single-spaced twin exists, so the series does not split. A lookup keyed on the single-spaced name returns zero rows instead. A second series, thick film high ohm, carries a trailing space on 488 parts in both Series and Series Group. For the structural problem with that array file, see the chip resistor array size column.

The rarest defects are the most precise. One MLCC row, 1808B102M252, has a tab character in the Size field, and three MLCC cells begin with a space: MT31N102F500 and 0201X224K160 in Tolerance, 1206B475M500 in Termination.

How many values split in two when whitespace is not trimmed?

Twenty-five, in nine columns. We took every distinct value in every column of 00_ALL_MLCC_parts_combined.xlsx and the other 119 files, collapsed each run of whitespace to one space, trimmed the ends, and counted how many values merged. 62 distinct strings in the MLCC L column collapse to 54, and 51 in W collapse to 43, so a group-by on either column returns 8 groups that should not exist. Nine columns across three families lose 25 values in all.

The MLCC Remark row connects to our post on the customised-part flag, which counted rows. We counted part numbers: 321 carry the flag, 299 with a line break and 22 with a space. A filter for the flag that expects either form misses the other one.

Why do 349 of 457 radio-frequency parts hold a line break?

Because the radio-frequency exports pack several bands into one cell, separated by a line break. We counted line breaks cell by cell. The RF family holds 1,483 line breaks in 479 cells, on 349 parts, and Walsin_Filter.xlsx alone has them on 214 of 235 parts, 91.1 percent. A cell such as Insertion Loss (dB) on RFCPL1608070P2T holds three readings, one per band, on three lines.

The same cells carry padding. The longest run in the catalogue is 170 consecutive spaces, in the Insertion Loss (dB) cell of RFLPF1608060Y1T, a cell of 191 characters. Six cells hold a run of ten spaces or more in the RF family and three in the antenna family, the longest antenna run being 67 spaces in the Peak Gain cell of RFDPA141300SBLB801. A line-based tool, or a text export that does not quote the cell, will split one RF part across several records. For what the radio-frequency columns can and cannot answer, see the 203-file census.

What should a Swedish import step do with these cells?

Three things, and the order matters. A cross-reference that compares text will fail in the way this post counts; one that normalises first will not. That is how our free cross-reference tool treats these columns, and it is why we state differences from your original rather than hiding them behind a tick.

1. Split on the line break before you trim. In the 479 RF cells the line break is the separator between bands. Collapse it to a space first and three readings become one unreadable string.
2. Then collapse every run of whitespace to one space and trim the ends. That removes the 25 phantom values and the 5,476 array names that a single-spaced lookup would miss.
3. Leave the Part Number column alone. It is clean in all eight families, and altering it breaks the join.

We hold 28 Walsin datasheets, 451 pages, covering 26 MLCC series and 2 chip resistor series, and the datasheet is the document to qualify against, not a trimmed export cell. For thin film, inductor, varistor, disc capacitor, radio-frequency and antenna parts we hold none, so on those lines the export is the only figure we can show you, and it should be confirmed on Walsin’s own site.

What is in Hong Kong is on ready stock: 1,125 part numbers as at 23 September 2026, a dated snapshot and not a live feed. The capacitor range sits on the MLCC page.

The short version

Whitespace hides in 7,424 cells across 6,682 part numbers in 5 of Walsin’s 8 families. Radio frequency is worst at 353 of 457 parts, 77.2 percent, with line breaks separating bands. Thin film, disc capacitor and varistor, 95,222 part numbers, hold none. Doubled spaces are the commonest defect at 5,620 cells, almost all in one chip resistor array column. Collapsing and trimming merges 25 phantom values in nine columns, 8 each in the MLCC L and W columns. The Part Number column is clean in every family. We will not tell you a value is the same as another until it has been trimmed. We will show you the string the file holds. Qualification stays with you.

Send us the descriptive columns off your Swedish BOM import and the groups they produced, and we will tell you which groups are one value in two. And a question back: when your last group-by on a length column returned eight more groups than the part has lengths, which row did you doubt first?

Frequently asked questions